As communication and information technology rockets now, and along with the maturation of Internet Protocol (IP) technology, the application of the Internet spreads rapidly. People are no longer satisfied with monotonous voice communications. Instead, they demand brand new multimedia communications. At the same time, mobile communication networks and fixed communication networks still serve as the main communication medium. Therefore, it is acknowledged in the industry that it is the trend that the mobile communication network and the fixed communication network will develop towards IP and the Internet will merge with the telecommunication network. In other words, an IP multimedia communication system which supports multiple access technologies is a focus of the future development in the industry. An IP Multimedia Subsystem (IMS) and a Next Generation Network (NGN) are just the kind of networks which support multiple access technologies while implementing IP multimedia applications.
The IMS is a network subsystem introduced by the 3rd Generation Partnership Project (3GPP) on the basis of a Universal Mobile Telecommunications System (UMTS) Packet Switched (PS) domain to achieve IP multimedia communication control. The IMS supports multiple access technologies, shields user access technology, controls the openness of service capability and provides IP multimedia communication experience based on customized user data. The IMS employs the UMTS PS domain or other IP access networks as a bearer for upper layer control signaling and media transmission. Major functional entities in the IMS include: a Call Session Control Function (CSCF) for controlling user registrations and sessions, an Application Server (AS) for providing various service logic control functions, a Home Subscriber Server (HSS) for centralized administration of subscription data, and a Multimedia Gateway Control Function/IM Multimedia Gateway Function (MGCF/IM-MGW) for interworking with a Circuit Switched (CS) network. In the IMS, a user accesses the IMS through a Proxy-CSCF (P-CSCF), which is a proxy node in a visited network where the user is currently located. While the session and service triggering control and the service control interaction with the AS are performed by a Service-CSCF (S-CSCF), which is a service node in a home network of the user.
The IMS provides IMS users with IP multimedia value-added services through various ASs. In the IMS, services are triggered based on an initial Filter Criteria (iFC) evaluation. The iFC is a major part of the user subscription data saved in the HSS and downloaded by the S-CSCE assigned to the user during registration. The iFCs with different priority levels defines different service triggering conditions and corresponding destination ASs. The S-CSCF compares a service request received from or sent to a user it serves with the service triggering conditions in the iFCs one by one according to their priorities. If the service request matches the service triggering condition in an iFC, the service request will be sent to the corresponding AS defined by this iFC. If the service request does not match the service triggering conditions in the iFCs, the S-CSCF continues the evaluation of the next iFC in the descending priority sequence. Or, if the service request matches the service triggering condition in an iFC and is sent to the corresponding AS and the AS returns the service request to the S-CSCF, the S-CSCF also continues the evaluation of the next iFC in the descending priority sequence. After finishing the evaluation of all the iFCs, the S-CSCF sends the service request to a next network node according to a destination identifier in the service request. In this method, services are triggered according to uniform filter criteria which are independent with specific service. Thus the processing model employed by the IMS service node S-CSCF can be shared and reused by various services, so that the IMS architecture is able to provide rich and various customized services through various ASs.
The NGN is a convergent network based on packet technologies, mainly adopting packet switched technologies and employing a bearer/control separated architecture. The NGN inherits all the services of the previous fixed network as well as some of the mobile network. The NGN integrates advantages of the fixed network, the mobile network and the IP network so as to allow users, such as analog users, digital users, mobile users, Asymmetrical Digital Subscriber Line (ADSL) users, Integrated Service Digital Network (ISDN) users, IP narrow-band network users, IP broadband network users and even satellite users, to communicate with each other in the NGN.
Both the IMS and the NGN adopt a Session Initiation Protocol (SIP) as their session control protocols. The SIP is one of the multimedia communication system frame protocols defined by Internet Engineering Task Force (IETF), and is an application layer protocol for the establishment, modification and termination of a multimedia session. Since the SIP is based on published Internet standards and is innately advantageous with respect to the combination and interworking of a voice service and a data service, it can implement session controls across media and equipment and support varieties of media formats. It should be noted that the SIP supports interaction between two participants of a session to exchange Session Description Protocol (SDP) descriptions of the media streams desired to be exchanged in an offer/answer mode during a session establishment, so as to accomplish a negotiation of the exchanged media streams in the session. In an established session, a participant can also exchange the SDP description to change the media stream desires to be exchanged by re-negotiation, i.e., using the offer/answer mode again, to dynamically add/delete media streams or modify the attribute of the exchanged media streams. In this way, it is easier to present abundant multimedia service features. At the same time, the SIP supports pushing the intelligence towards applications and terminals and so as to relieve the burden of the network. The SIP also supports application layer mobility capability including dynamic registration, location management, and re-direction mechanisms, as well as the Presence capability (determining the specific communication means according to the location and status of the user), Fork capability (an ability to perform serial attempt or parallel forwarding a service request to multiple valid contact address registered to a destination identifier), and Subscribe and Notify mechanism, which makes it facilitated for the deployment of new services. Moreover, the SIP is a simple protocol with recognized extension potential and is thus widely applied in networks including the IMS and the NGN.
In the IMS and the NGN, a user can access the network through a multi-mode terminal, or through different terminals, via access networks of different access technologies, and obtain uniform multimedia services according to his/her subscription.
In the IMS and the NGN, a user binds his/her current terminal contact address and service identifier (which is the user's IP Multimedia Public identity (IMPU) in the IMS) through registration. Therefore, during a registration period, there should be a unique access address corresponding to the service identifier of the user. In order words, once the access address changes, the user needs a re-registration, and the original access address will be deleted, and the on-going session will be released if no special processing is performed in the service layer. Therefore, when a user in a session needs to change his/her Access Point (AP) or access technology which further result in the change of the access address, or even change the terminal, the problem of maintaining session continuity emerges.
On the other hand, a voice call to a peer user (e.g., a subscriber of the IMS or the NGN) can be either established as a Voice over IP (VoIP) call using an end-to-end IP bearer under the control of the IMS/NGN, or directly established as an interworking call between traditional CS domain and the IMS/NGN. Some access technologies through which the terminal accesses the IMS or the NGN belong to hot spot coverage technologies (e.g., Wireless Local Access Network (WLAN)). When a user accesses IMS/NGN through the WLAN to initiate a VoIP call, if the user roams out of the hot spot coverage of the WLAN, the network connection will be lost. At this time, if the user can change to access the CS domain through Wideband Code Division Multiple Access (WCDMA) which usually implemented with wide, continuous coverage to continue the session with the peer user, the session continuity will be well kept. Similarly, a WCDMA system, which is designed to cover wide areas continuously, may fail to provide high quality wireless connection in some buildings due to penetration loss of wireless signals, while hot spot coverage technologies, represented by the WLAN, are able to fill the coverage holes emerged here. In other words, if the user changes a CS voice call to an IMS VoIP session through the WLAN access when he moves into such a coverage hole of WCDMA, the session continuity will be well kept. Considering that the call between the user and the peer IMS/NGN user through the CS domain actually includes a CS call connection from the user to an interworking gateway between the CS domain and the IMS/NGN, and a SIP session from the interworking gateway between the CS domain and the IMS/NGN to the called IMS/NGN user. A CS call, which is converted to an IMS/NGN session through the interworking gateway, can be regarded as a special access means to the IMS/NGN from the point of view of the IMS/NGN. So the problem of maintaining session continuity when a terminal changes access means in an IMS/NGN session should also be taken into consideration to better guarantee the session continuity.
There exists a possible solution to guarantee the session continuity, i.e., Unlicensed Mobile Access (UMA).
The UMA standards are a set of published specifications jointly developed by a number of leading operators and vendors within the wireless industry. The UMA standards enable a terminal to access a cellular network and acquire services of the cellular network through unlicensed spectrum access technologies such as WLAN and Bluetooth (BT), and eventually enable a dual-mode terminal to roam and handover between the cellular network and the unlicensed spectrum wireless network.
FIG. 1 is a schematic diagram illustrating a process of maintaining session continuity through the UMA standards according to the prior art. As shown in FIG. 1, a UMA network controller (UNC), whose network location is similar to that of a Base Station Controller (BSC) or a Radio Network Controller (RNC) in the cellular network, is added in advance into the unlicensed spectrum wireless network, e.g., the WLAN. The added UNC is used for providing an interface with the core network just like that provided by the BSC/RNC in cellular network. Therefore the core network may regard the unlicensed spectrum wireless network as a normal cellular access network. When a terminal roams and handovers between the cellular network and the unlicensed spectrum wireless network, handover between the two networks are implemented through the interaction among the core network, the UNC in the unlicensed spectrum wireless networks and a corresponding entity, such as the BSC, in the cellular network.